OPTICAL ILLUSION IMAGES DATASET

UDC 612.821.89:004.8

Robert Max Williams and Roman V. Yampolskiy

University of Louisville
Louisville, Kentucky, United States

Author’s contact information: robertmaxwilliams@gmail.com
Author’s contact information: roman.yampolskiy@louisville.edu

INSAM Journal of Contemporary Music, Art and Technology, Issue 2, 2019

Main Theme of the Issue: Artificial Intelligence in Music, Arts and Theory

Publisher: INSAM Institute for Contemporary Artistic Music, Sarajevo, Bosnia and Herzegovina

Section: THE MAIN THEME

Abstract: Human vision is capable of performing many tasks not optimized for during its long evolution. Reading text and identifying artificial objects such as road signs are both tasks that mammalian brains never encountered in the wild but are very easy for us to perform. However, humans have discovered many very specific tricks or illusions that cause us to misjudge the color, size, alignment, and movement of what we are looking at. A better understanding of these phenomenon could reveal insights into how human perception achieves these extraordinary feats. In this paper we present a dataset of 6,725 illusion images gathered from two websites, and a smaller dataset of 500 hand-picked images. We will discuss the process of collecting this data, models trained on the data, and the work that needs to be done to make this information of value to computer vision researchers.

Keywords: Computer Vision, Optical Illusions, Human Vision, Machine Learning, Neural Networks, Cognition

PDF:

This image has an empty alt attribute; its file name is INSAM-naslovna-.jpg
ISSN 2637 – 1898
On the cover: Devine Lu Linvega / NASA, The Puppyslug Nebula, courtesy of NASA and Google DeepDream
 Design and layout: Milan Šuput, Bojana Radovanović